Probably yes, but why? After all, the world has had millions of victims of extreme misery for some time now.
What about opportunity cost? Why should you aid these people and not others? I can think of two arguments:
1. Aid is more effective when large numbers of donors coordinate upon addressing a single disaster in a focused manner.
2. Donors are frail in their commitment to altruistic ends. A dramatic headline induces them to give when they would not otherwise be generous. So if you find that images of tsunami victims tug at your heartstrings, give now because you will forget about being kind a week or month from now.
These two hypotheses are not fully distinct. Even if #2 does not describe you, it probably applies to many others. If your donations help make the tsunami victims a more "focal cause" this will induce more giving from altruistically frail others.
But when are these economies of scale exhausted? I doubt if the generated aid will come close to saving all the relevant lives. If these donations do in fact produce increasing returns across a giving network, why are you giving to any other causes?
Is it that donations will stop at the point when saving additional Sri Lankan lives hits a steeply upward-sloping cost curve? (Unlikely.) Or are we so altruistically frail that if we displace our other giving, we will have fewer giving motives on net and thus will give less overall? Can we not sit down and understand our frailty rationally and write one big check, once a year, to the "most efficient cause"? Possibly not.
Thanks to Jonathan S. for the pointer. And here is my previous post on whether you should give money to beggars.